create_synthetic_forecasting_dataset#

openstef_core.testing.create_synthetic_forecasting_dataset(start: datetime = datetime.fromisoformat('2025-01-01T00:00:00+00:00'), length: timedelta = timedelta(days=30 * 9), sample_interval: timedelta = timedelta(hours=1), random_seed: int = 42, wind_influence: float | None = -0.2, temp_influence: float | None = 0.3, radiation_influence: float | None = -0.2, stochastic_influence: float | None = 0.1, other_components: dict[str, float] | None = None, *, include_atmosphere: bool = False, include_price: bool = False, include_available_at: bool = False) → TimeSeriesDataset[source]#

Create synthetic forecasting dataset for testing.

Generates time series data with configurable components influencing load.

Parameters:

  • start (datetime) – Start datetime for the dataset.

  • length (timedelta) – Total duration of the dataset.

  • sample_interval (timedelta) – Time interval between consecutive samples.

  • random_seed (int) – Random seed for reproducible random components.

  • wind_influence (float | None) – Coefficient for wind speed component on load.

  • temp_influence (float | None) – Coefficient for temperature component on load.

  • radiation_influence (float | None) – Coefficient for radiation component on load.

  • stochastic_influence (float | None) – Coefficient for random noise component.

  • other_components (dict[str, float] | None) – Additional components with their influence coefficients.

  • include_atmosphere (bool) – Add pressure (~1013) and relative_humidity (~70%) columns.

  • include_price (bool) – Add day_ahead_electricity_price (~50) column.

  • include_available_at (bool) – Add available_at column (index + sample_interval).

  • start

  • length

  • sample_interval

  • random_seed

  • wind_influence

  • temp_influence

  • radiation_influence

  • stochastic_influence

  • other_components

  • include_atmosphere

  • include_price

  • include_available_at

Returns:

TimeSeriesDataset containing synthetic load and component data.

Return type:

TimeSeriesDataset